Table 2.

Model comparison on gp120_Benchmark_1 dataset

Models	MSERP	MSESP	MSELTNP	MSEOverall	CCI	STD	TBM
SV-HMM	38.02	30.46	45.57	30.21	67.82	0.8	2.62
HMESVM	36.35	24.13	44.43	29.76	69.38	1.9	1.35
SVMLIB	54.53	20.68	44.43	32.65	67.39	1.1	0.17
DecorateSVM	81.82	10.34	55.55	32.88	65.31	2.7	3.02
MLP	72.43	27.59	55.57	27.27	57.14	4.2	4.14
RBFN	63.63	27.58	77.78	29.32	55.10	3.8	0.06
Logistic	63.63	34.48	66.66	31.35	53.06	1.0	0.11
J48	90.91	37.93	66.67	37.77	44.90	3.0	0.01

The final CCI value was calculated based on the average of all the prediction accuracies ob-served during the tenfold validation process. MSE, TBM and STD (ST Deviation obtained from ten sub-samples) indicate the accuracy, complexity and stability of the model respectively.